I have worked with several face-centered cubic alloys and metal hydrides lately where there are subtle but clear tendencies of the two first peaks, 111 and 200, to be shifted towards each other. In other words, the 111 peak is at slightly too high scattering angle and the 200 peak is at slightly too low angle compared to the calculated profile in Rietveld refinements. In addition, there is some anisotropic line broadening.
According to Warren's "X-ray diffraction" (page 288-290,
http://docshare04.docshare.tips/files/28270/282700659.pdf), Makinson et al. (2000) (
https://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.296.2371&rep=rep1&type=pdf) and other, such peak shifts are a signature for stacking faults in the fcc structure.
I have fitted models with stacking faults to my data, with John Evens tutorial on Cu (
http://community.dur.ac.uk/john.evans/topas_workshop/data/cu_tutorial_01.inp) as a starting point. I get a good description of the anisotropic line broadening, typically with around 5-10% stacking faults (pa = 0.90-0.95), but the shifts in peak positions are not modeled. I've also examined simulated patterns with different stacking fault probabilities without seeing any peak shifts. According to Warren's analytical expressions, a stacking fault probability of 5% should give a pretty significant shift of almost 0.2 degrees in 2theta (assuming a_fcc 3.5 Å and Cu-radiation).
Have anybody found a way to model peak shifts from stacking faults, as described by Warren, with Topas 6 while maintaining a good description of the peak broadening?